The next step was to count and measure the tree rings.  A microscope was used to enlarge the samples.  The tree rings were measured using the calipers on the eyepiece at 2X.  The ring width measurements were taken to 0.05 of a millimeter. 



Methods: Data Analysis

              The raw tree ring width measurements were entered into an Excel spreadsheet.  For each tree sample, two cores were taken.  The measurements of each core within a sample were averaged together to give an average ring width.  The raw data attained in the average ring widths were then plotted onto a time-series graph.  The data needed to be corrected to portray an accurate measurement.  This process is called standardization.  The standardization eliminates age related growth patterns in a tree. An exponential regression line was fit into the data points.  The resulting exponential regression equation was then used to standardize the raw data.  Shown by equation 1:
                       I_t = R_t / G_t                        (eq. 1)
    I_t       -  standardized dimensionless unit
    R_t     -   raw ring width
    G_t    -  expected ring width
                                                                      (Cook, Briffa, Shiyatov, Mazepa, 1989)

The standardized data is expressed in dimensionless units where 1.00 is the normal growth for every data point.   
               Climatological data was obtained from the NCDC (National Climatic Data Center) and NOAA (National Oceanic and Atmospheric Administration).  The types of climatological data were precipitation, average temperatures, maximum temperatures, and minimum temperatures.  The data obtained was for the years 1930-1999.  The data that was used was a collection of latitude 39.13N and longitude 83.62W (maximum and minimum temperatures), and a compilation of Southwest Ohio data (precipitation and average temperatures).  The Average temperature and precipitation data were obtained from NOAA via the NOAAServer.  The maximum and minimum temperature data was collected from the NCDC.  The data was in the form of monthly averages.  The data was then sorted to show only the months between April and September, due to the length of Oxford's growing season.  This was put combined with the tree sample data in an Excel spreadsheet to produce a final data sheet of five sample standardized tree ring widths, precipitation, average temperatures, maximum and minimum temperatures for the study.
               The standardized data from each sample was then grouped into decades.  For each sample the decades were analyzed to see if there were changes of tree growth from decade to decade.  In the Excel spreadsheet, a two-tailed student t-test was used for the analysis.  The alpha was set to 0.05.  The t-test was one assuming equal variances.  The null hypothesis was that the two data sets were equal.  For each sample, two chronological decades were analyzed (e.g. 1990s and 1980s; 1980s and 1970s; etc.).  This test was repeated for each decade. 
               Using Statview 5.0, the Spearman Rank Correlation was used to analyze the data.  The test was used to analyze the correlation between the samples in five-year increments.  The alpha was set to 0.1. The null hypothesis was that the two data sets were independent of each other.  Averages of ring width for five years were taken from each sample.  Each sample was compared with the other samples.  The Spearman Rank Correlation was also used to test the sample and the climatological data.  The data for this test was divided into one-year increments to give a higher resolution.  Each sample was tested with the other samples to find trends in yearly growth.  Also, each sample was tested with the precipitation, average temperature, and maximum and minimum temperature yearly data.  This procedure was repeated for each sample.

Results:
Student T-test Assuming Equal Variances:

Sample 1:  Between the decade of the 1990s and 1980s the p-value was 0.04.  Between the decade of the 1980s and 1970s the p-value was 0.30.  Between the decade of the 1970s and 1960s the p-value was 0.005.  Between the decade of the 1960s and 1950s the p-value was 0.88.  Between the decade of the 1950s and 1940s the p-value was 0.03.  Between the decade of the 1940s and 1930s the p-value was 0.34.
Sample 2:  Between the decade of the 1990s and 1980s the p-value was 0.25.  Between the decade of the 1980s and 1970s the p-value was 0.55.  Between the decade of the 1970s and 1960s the p-value was 0.79.  Between the decade of the 1960s and